Comparative risk of nonvertebral fractures associated with oral anticoagulants in patients with venous thromboembolism

Oral anticoagulants are widely used to prevent recurrence in patients with venous thromboembolism (VTE), but studies of these drugs in patients with atrial fibrillation have raised concerns that warfarin may be associated with higher risk of fractures than direct oral anticoagulants. Using the United States administrative claims data from Medicare and Optum Clinformatics (January 2016 to June 2024), we conducted a cohort study that emulated a three-arm target trial of VTE patients who initiated apixaban (N = 73 668), rivaroxaban (N = 30 849), or warfarin (N = 20 872). The primary outcome was nonvertebral fractures. Propensity score matching weights were used to balance baseline covariates. Patients were followed while on treatment; intention-to-treat (ITT) analyses were also conducted. Weighted hazard ratios while on treatment were 1.12 (95% CI, 0.89-1.41) for apixaban vs. warfarin, 1.13 (95% CI, 0.87-1.46) for rivaroxaban vs. warfarin, and 0.99 (95% CI, 0.80-1.23) for apixaban vs rivaroxaban. Results were consistent for ITT and other sensitivity analyses and across subgroups of age, sex, osteoporosis, and chronic kidney disease. Our findings suggest that apixaban, rivaroxaban, and warfarin are comparable with regards to nonvertebral fracture risk when used for VTE secondary prevention.